Validation of bacterial endotoxin test for injectable batch produced cytotoxic drugs
3 October 2013P.H. Secrétan, MN-G. Guerrault-Moro, R. Vazquez, S. Crauste-Manciet CHI Poissy Saint Germain-en-Laye, 20 rue Armagis,
Saint Germain-en-Laye, 78105. France
Evaluation of the feasibility of bacterial endotoxin test (BET) as a part of the final control for batch release of standard doses of cytotoxic drugs.
An automated photometric BET was selected that used dried, precalibrated Limulus amebocyte lysate cartridges (Endosafe®-MCS, Charles River®). Maximum limit dilutions were determined for each cytotoxic solution in accordance to endotoxin limits given in American pharmacopeia (USP). BET conditions were verified in triplicated on the lowest sensitivity 0.005 EU/mL cartridges by measuring the recovery of endotoxin positive controls in sterile water dilutions for 6 cytotoxic standard solutions [cyclophosphamide (CP), docetaxel (DTX), 5 fluorouracil (5FU), gemcitabine (GZ), irinotecan (IRI), oxaliplatine (0xPt)] and 2 diluants (Dextrose 5% (D5), Sodium Chloride (NS) 0.9%)]. Dilutions of injectable cytotoxic solutions were performed using endotoxin-free vials and endotoxin-free water for injection (WFI) or endotoxin-free buffer (BG 120 Charles River®). Acceptance criteria was spike recovery of the positive control within the BET acceptance range of 50%-200% and CVs for reaction times for both the samples and positive controls replicates less than 25%.
For each injectable solution, the maximum dilution limit was calculated in accordance to USP, results are given in following table.
D 5% needed 1:50 dilution to detect control endotoxin. Inhibition can be explained by acidic pH of the dextrose solution (3.77), no dilution was required for NS 0.9%. A systematic 1:10 dilution in endotoxin free WFI for all cytotoxic drug solutions was sufficient to detect positive control endotoxin except for docetaxel. A 1:40 dilution with specific BG 120 buffer was required to counteract the inhibiting effect of surfactant included in the docetaxel formulation. In most of the cases, 1:10 dilution allowed detecting a potential contamination with endotoxin above 0.05 EU/mL which in all cases 10 fold below the pharmacopeial endotoxin limit of 0.5 EU/mL. For docetaxel, 1:40 dilution gives sensitivity of 0.2 EU/mL which is still below pharmacopeial endotoxin limits but just at the limit maximum dilution.
BET for cytotoxic solution using automated photometric method was given quick results in 20 minutes and able to detect pyrogen contamination in cytotoxic preparation. It can be easily included in the control release process of the batch as a complement method of sterility testing.